Complex networks have gained more and more attention since the end of the ‘90s. Across a wide variety of fields, a lot of data has been collected, that can be represented by networks. Clustering is a widely used tool of network analysis, that can help us acquire valuable information about these complex systems. The main stumbling block for cluster analysis is often the large size of the networks. In order to overcome this obstacle, new and efficient methods are needed. In this work, I will address this problem. I will introduce a novel method for examining technological changes based on the United States patent citation network, then I will propose a novel, efficient, and accurate algorithm for graph clustering. Furthermore, I will present and illustrate a novel framework for describing the cluster structure of networks.